Quantum Chemistry and Physics

We seek to develop novel quantum simulation techniques that leverage principles from physics, chemistry, applied mathematics, and data science to address outstanding questions in quantum chemistry and physics. We are best known for our development of stochastic methods (i.e., Quantum Monte Carlo methods), which use random numbers to perform highly accurate electronic structure calculations. Topics of recurring interest include unraveling the physics of 2D materials and their heterostructures, exotic forms of magnetism, and entangled states. Examples follow below:

Finite Temperature Electronic Structure

Our group has developed a variety of stochastic finite temperature electronic structure methods that can be used to describe models and materials with small bandgaps or at high temperatures and/or pressures. These many-body methods have been shown to accurately predict finite temperature crossovers and phase ordering. More recently, we have developed new, more efficient means of performing these simulations in the canonical ensemble, which opens doors to the more accurate study of nuclear matter and certain cold atom experiments, and new ways of computing the entanglement of fully interacting systems.

Funding: NSF CAREER Award

Low-Dimensional Materials

The discovery and engineering of 2D materials, including graphene, boron nitride, and the transition metal dichalcogenides, has enabled researchers to develop a wide range of novel devices that exhibit novel phenomena in a relatively controlled setting. Our group has been leveraging high-accuracy quantum chemistry and stochastic methods to illuminate the physics within these materials with accuracies not readily attainable using other approaches. In particular, we have developed new means of modeling magnetic and multilayer graphene heterostructures that have revealed novel types of order and collaborated with experimentalists to realize and understand these orders in real quantum materials. Our approaches are unique in their ability to account for strong correlation, yet scale to the system sizes need to reveal emergent physics.

Funding: DOE Center for the Predictive Simulation of Functional Materials and Camille Dreyfus Teacher-Scholar Award

Correlated Catalysis

Our group is developing a variety of tools to model heterogeneous catalytic reactions on correlated materials. We have shown that modeling such reactions is well within the range of stochastic methods and that we can accelerate these methods using machine learning and other surrogate approaches to yield high-accuracy ground state geometries and transition states.

Funding: Air Force Office of Scientific Research